How LinkedIn Performs Maintenances at Scale
summary
LinkedIn runs hundreds of thousands of servers across global data centers and uses impact analysis to map affected hosts and services for a maintenance (notably single top‑of‑rack switch upgrades). Services provide either simple static headroom rules (e.g. ≤5% instances down) or custom REST API checks; approvals are all‑or‑nothing and refresh periodically. An approval management queue coordinates safe batches and a controller triggers automated execution to gracefully drain, upgrade, and restore — shrinking a weeklong switch maintenance to under an hour.
AI-written summary · about
chapters (6)
related talks
29mAdapt to Unified and Pluggable Cluster Management Platform at LinkedIn
31mAutomated Operating System and Environment Certification at LinkedIn...
29mCommas Save Lives, or at Least LinkedIn
41m6 years of Gradle at LinkedIn
20mReliable Migration Of 4+ Million CPU Cores of Compute to Kubernetes
28mDesigning for Failure: How to Manage Thousands of Hosts Through Automation
19mDelete This: Decommissioning Servers at Scale
06mLightning Talk: Keeping the Show Running: Backstage Maintenance from a Platform Eng...